Score any startup idea in 60 seconds. No API keys. No setup.
When a user says /lobstr "idea", you run a full startup idea scan:
- Competitor search — finds real competitors via live web search
- LOBSTR score — scores the idea across 6 dimensions (0–100 each)
- GRID investor signal — checks how many EU VCs are active in this space
You get back a formatted score card to show the user. By default, output is private — nothing is published anywhere.
Use lobstr whenever a user:
- Asks you to evaluate, validate, or score a startup idea
- Wants to know if a business concept is worth pursuing
- Needs a competitive landscape analysis for a new product
- Asks "should I build this?"
# Default: private score card
python3 scripts/lobstr.py "the startup idea goes here"
# Publish to runlobstr.com and get a share URL
python3 scripts/lobstr.py "the idea" --public
# Also post to m/lobstrscore on Moltbook
python3 scripts/lobstr.py "the idea" --public --moltbook
# Raw JSON output (for agent-to-agent piping)
python3 scripts/lobstr.py "the idea" --jsonNo API keys are needed — the skill calls the hosted scoring API at runlobstr.com automatically.
| Flag | Effect |
|---|---|
| (none) | Private output — score card to stdout only |
--public |
Publish to runlobstr.com and show share URL |
--moltbook |
Post to m/lobstrscore on Moltbook |
--json |
Output raw JSON instead of formatted card |
🦞 LOBSTR SCAN
"AI-powered legal contracts for SMEs"
LOBSTR SCORE 45/100 [====------]
L Landscape 🔴 15/100 Saturated with well-funded players.
O Opportunity 🟢 75/100 Large market, every SME needs contracts.
B Biz model 🟡 60/100 SaaS model is solid, defensibility questionable.
S Sharpness 🔴 20/100 Undifferentiated. AI is table stakes.
T Timing 🟢 85/100 Perfect timing. Legal tech is hot.
R Reach 🟢 70/100 Good reach via online channels.
VERDICT
Tough market, undifferentiated product. Needs a radical rethink.
🚫 NOT YET.
| Dimension | What it measures | |
|---|---|---|
| L | Landscape | How crowded is the competitive space? |
| O | Opportunity | Is there a real, large market? |
| B | Business model | Is monetization clear and defensible? |
| S | Sharpness | How differentiated vs alternatives? |
| T | Timing | Is the market timing right? |
| R | Reach | How easily can this scale? |
Score ranges: 🟢 ≥ 70, 🟡 50–69, 🔴 < 50. Overall score is weighted (S and B count more).
The default mode uses the free hosted API (5 scans/day, no keys needed).
For unlimited scans, the user can set both:
export ANTHROPIC_API_KEY=<key>
export EXA_API_KEY=<key>This activates BYOK mode — the skill runs the full pipeline locally instead of calling runlobstr.com.
ClawHub: https://clawhub.ai/rednix/lobstr
- runlobstr.com — score cards and scan archive
- GRID — EU investor database
- NMA — built by NMA
MIT